Rolling Stone’s 100 Greatest

Using Spotify and Billboard to measure the endurance of Rolling Stone’s 100 Greatest Artists

Course code: MY472 Candidate number: 25710 Word count: 742

The GitHub repository for this project can be found here.

Introduction

Released in 2010, the Top 100 Greatest Artists list represents what Rolling Stone magazine considers to be the best of the best when it comes to music.

13 years later, this project analyses the ranked artists and their music to see how they stand in 2023. Are they still popular… or have they faded into obscurity? Are there any specific characteristics that seem to explain enduring engagement?

Research Question:

## [1] "Rolling Stone Magazine ranked their 100 greatest musical artists of all time. At the end of 2023, how has their music endured? Are there any features or characteristics that seem to explain enduring engagement?"

Approach:

As claimed by Matt Daniel in his project on “The most timeless songs of all-time”, “Popularity on Spotify … is a strong signal for whether a song will be remembered by our children’s children.”

Taking this into account, this project uses Spotify’s Artist Popularity, which is measured from the popularity of all the artist’s tracks, as a proxy for enduring engagement with an artist’s music.

However, recognizing that there might be factors that affect current popularity but not endurance such as the reappearance of songs in pop culture, another proxy for endurance is used; the number of weeks an artist has been on the Billboard Artist 100 chart, which denotes historical popularity.

Using these two factors - current and historical popularity - this project builds an endurance index which examines how enduring the Rolling Stone’s 100 Greatest Artists and their music is in 2023.

## [1] "Artist Billboard Popularity = Normalized Weeks on Chart = (Weeks on Chart / Max Weeks on Chart (495)) * 100"

## [1] "Endurance Index = (Artist Spotify Popularity + Artist Billboard Popularity) / 2"

From this endurance index of the artists, the project then looks at the characteristics of the artists top track to see if there are any features that explain enduring engagement, such as genre, decade, or mood.

Data

A summary of the data sources and variables collected is shown in Table 1 and the list of variables used in the final database is shown in Table 2.

Table 1: Data Sources and Variables Collected

Data Source	Access Method	Processing Method
Rolling Stone 100 Greatest rankings	Rolling Stone Magazine Website	Webscraping
Spotify	Spotify Web API	API calls
Billboard Artist 100	Billboard Website	Webscraping

Table 2: Final Database Variables

Variable	Source	Explanation
Artist Name	Rolling Stone 100 Greatest rankings
Rolling Stone Ranking	Rolling Stone 100 Greatest rankings
Current Popularity	Spotify	Artist popularity on Spotify
Historical Popularity	Billboard	Weeks on Chart in Artists 100 Chart
Endurance Index	Billboard & Spotify
Genre	Spotify	Simplified genre of artist (eg. rock instead of southern rock)
Track	Spotify	Top Track of each artist
Decade	Spotify	Release Decade of Track
Danceability	Spotify	Danceability score of Track
Energy	Spotify	Energy score of Track
Happiness	Spotify	Valence score of Track
Mood	Spotify	Average of danceability, energy and valence score

Rolling Stone’s 100 Greatest Artists Playlist

Analysis

Ranking

Figure 1 shows that Rolling Stone’s 100 greatest artists, at the end of 2023, seem to have endured with an above average current popularity of 66 and 90% of artists with a score above 50. However, Figures 2-3, which account for historical popularity, demonstrate that if we look at the popularity of these artists from 2014 (inception of Artist 100 Billboard chart) to now, the 100 greatest do not appear to stand the test of time as the average endurance index is 37 and 56% of artists having an historical popularity score of 0.

When looking at ranking as a potential determinant of endurance, the relationship between all endurance variables is negligible (below ± 0.199 correlation) suggesting ranking does not explain endurance.

Current Popularity

Historical Popularity

Endurance Index

Characteristics of Enduring Artists: Genre, Decade and Mood

Genre

Hip hop appears to be the most enduring genre and pop the least. However, this analysis cannot answer the question of whether music of certain genres is more enduring due to a limited and skewed sample (e.g. reggae has only one artist and rock artists account for 40% of artists).

Decade

The most enduring decades appear to be the 90s, the 80s and the 2000s and there is a weak positive correlation between decade and endurance which suggests the more recent the music the more enduring. However, it must be noted the data is skewed with 65% of the music being from the 1960s-1980s and that this analysis is only performed to the top track of the artists.

Mood Analysis

The mood analysis shows that the relationship between energy, happiness and overall mood and endurance is negligible as seen by Figures 7-9 and the correlation coefficient. Danceability, on the other hand, appears to have a weak positive correlation with endurance suggesting danceable music may be more enduring.

Danceability

Energy

Happiness

Mood

Conclusion

In conclusion, the music of the 100 Greatest has endured well if we look at current popularity but fairs worse when considering popularity over a longer period of time. Hip hop 80s-00s danceable music seems to have endured the best. Accordingly, the most enduring artist is 00s hip hop artist Eminem who has a danceability score of 91%.

Code Appendix

print("Rolling Stone Magazine ranked their 100 greatest musical artists of all time. At the end of 2023, how has their music endured? Are there any features or characteristics that seem to explain enduring engagement?")

print("Artist Billboard Popularity = Normalized Weeks on Chart = (Weeks on Chart / Max Weeks on Chart (495)) * 100")
print("Endurance Index = (Artist Spotify Popularity + Artist Billboard Popularity) / 2")

# Scrape rolling stone website to get table with ranking and artist name
library(RSelenium)

rD <- rsDriver(browser = c("firefox"), verbose = FALSE, port = netstat::free_port(random = TRUE), chromever = NULL)
driver <- rD$client

url <- "https://www.rollingstone.com/music/music-lists/100-greatest-artists-147446/"
driver$navigate(url)

# Locate the reject button:
reject_button <- driver$findElement(using = "xpath", 
                                    value = '//*[@id="onetrust-reject-all-handler"]')
# Click on the button:
reject_button$clickElement()


# Artists dataframe with two columns: Ranking and Artist_Name and 100 rows
rolling_stone_artists_df <- data.frame(Ranking = numeric(100), Artist_Name = character(100), stringsAsFactors = FALSE)


# Click the load more button
load_more <- function(){
  # Load more button
  load_more_button <- driver$findElement(using = "xpath", 
                                      value = '/html/body/div[5]/main/div[2]/div[1]/div/article/div[3]/div[2]/div[2]/a')
  # Click on the button:
  load_more_button$clickElement()
}

# Click the load previous button
load_previous <- function(){
  # Load previous button
  load_previous_button <- driver$findElement(using = "xpath", 
                                         value = '/html/body/div[5]/main/div[2]/div[1]/div/article/div[3]/div[2]/div[1]')
  # Click on the button:
  load_previous_button$clickElement()
}


# Get the artist rankings
rank <- function(){
  # Find all elements with the class name "c-gallery-vertical-album__number"
  artist_rank <- driver$findElements(using = "class name", value = "c-gallery-vertical-album__number")
  
  # Extract text from each element
  artist_ranks <- sapply(artist_rank, function(element) element$getElementText()[[1]])
  
  # Print or use the extracted information
  return(artist_ranks)
}


# Get the artist names
name <- function(){
  # Find all elements with the class name "c-gallery-vertical-album__title"
  artist_name <- driver$findElements(using = "class name", value = "c-gallery-vertical-album__title")
  
  # Extract text from each element
  artist_names <- sapply(artist_name, function(element) element$getElementText()[[1]])
  
  # Print or use the extracted information
  return(artist_names)
}

# Scrape all 
scrape_artists <- function(){
  # Assign values from the vector to the first 50 rows of the "Ranking" column in top_hundred_artists_df
  rolling_stone_artists_df$Ranking[1:50] <- rank()
  rolling_stone_artists_df$Artist_Name[1:50] <- name()

  # Click the load more button
  load_more()
  Sys.sleep(2)

  # Assign values from the vector to the last 50 rows of the "Ranking" column in top_hundred_artists_df
  rolling_stone_artists_df$Ranking[51:100] <- rank()
  rolling_stone_artists_df$Artist_Name[51:100] <- name()
  
  return(rolling_stone_artists_df)
}

# Call function to scrape all
rolling_stone_artists_df <- scrape_artists()

# Clean artists names by changing Parliament and Funkadelic to Parliament
rolling_stone_artists_df$Artist_Name[rolling_stone_artists_df$Artist_Name == "Parliament and Funkadelic"] <- "Parliament"
rolling_stone_artists_df

# Close the RSelenium processes:
driver$close()

# Getrolling stone artists dataframe to csv

write.csv(rolling_stone_artists_df, "rolling_stone_artists_df.csv", row.names = FALSE)
# Access spotify API 
readRenviron("spotify-api.env")

# Get the client ID and secret (key) from the environment variables
client_id <- Sys.getenv("SPOTIFY_CLIENT_ID")
client_secret <-Sys.getenv("SPOTIFY_CLIENT_SECRET")

# Authenticate with the API 
library(httr)
authentication <- POST(
  'https://accounts.spotify.com/api/token',
  accept_json(),
  authenticate(client_id, client_secret),
  body = list(grant_type = 'client_credentials'),
  encode = 'form',
  verbose()
)
mytoken = content(authentication)$access_token
HeaderValue = paste0('Bearer ', mytoken)
# Get the artist id for each artist in the rolling_stone_artists_df data frame
library(jsonlite)

# Function to get the Spotify Artist ID for a given artist name
get_artist_id <- function(artist_name) {
  # Define the Spotify API endpoint for searching an artist
  search_url <- 'https://api.spotify.com/v1/search'
  
  # Set up the request with the access token
  search_response <- GET(
    search_url,
    query = list(q = artist_name, type = 'artist'),
    add_headers(Authorization = HeaderValue)
  )
  
  # Extract the artist ID from the response
  search_results <- content(search_response)
  
  # Check if any results were returned
  if (length(search_results$artists$items) > 0) {
    artist_id <- search_results$artists$items[[1]]$id
    return(artist_id)
  } else {
    # Return NA or any other value to indicate no match
    return(NA)
  }
}

# Create new dataframe based on rolling_stone_artists_df
spotify_artists_df <- rolling_stone_artists_df

# Apply the function to the entire "Artist_Name" column in the data frame
spotify_artists_df$Spotify_Artist_ID <- sapply(spotify_artists_df$Artist_Name, get_artist_id)


# Get the artist info for each artist in the top_hundred_artists data frame
get_artist_info <- function(artist_id){
  # Define the Spotify API endpoint for getting information about an artist
  artist_url <- paste0('https://api.spotify.com/v1/artists/', artist_id)
  
  # Set up the request with the access token
  artist_response <- GET(artist_url, add_headers(Authorization = HeaderValue))
  
  # Extract the artist information from the response
  artist_info <- content(artist_response)
  
  popularity <- artist_info$popularity
  
  # Create a list with the extracted information
  artist_data <- list(popularity = popularity)
  
  return(artist_data)
}

# Apply the function to the entire "Artist_Name" column in the data frame
result <- lapply(spotify_artists_df$Spotify_Artist_ID, get_artist_info)

# Extract individual elements
spotify_artists_df$Spotify_Popularity <- sapply(result, function(x) x$popularity)

spotify_artists_df

# add genre column
get_artist_genre <- function(artist_id){
  # Define the Spotify API endpoint for getting information about an artist
  artist_url <- paste0('https://api.spotify.com/v1/artists/', artist_id)
  
  # Set up the request with the access token
  artist_response <- GET(artist_url, add_headers(Authorization = HeaderValue))
  
  # Extract the artist information from the response
  artist_info <- content(artist_response)
  
  genre <- artist_info$genres
  
  # Create a list with the extracted information
  artist_data <- list(genre = genre)
  
  return(artist_data)
}

# Apply the function to the entire "Artist_Name" column in the data frame
result <- lapply(spotify_artists_df$Spotify_Artist_ID, get_artist_genre)

# Extract individual elements
spotify_artists_df$Genre <- sapply(result, function(x) x$genre)

spotify_artists_df$Genre <- sapply(spotify_artists_df$Genre, function(x) paste(x, collapse = ","))

spotify_artists_df$Genre <- ifelse(grepl("reggae", spotify_artists_df$Genre), "reggae", spotify_artists_df$Genre)
spotify_artists_df$Genre <- ifelse(grepl("soul", spotify_artists_df$Genre), "soul", spotify_artists_df$Genre)
spotify_artists_df$Genre <- ifelse(grepl("blues", spotify_artists_df$Genre), "blues", spotify_artists_df$Genre)
spotify_artists_df$Genre <- ifelse(grepl("folk", spotify_artists_df$Genre), "folk", spotify_artists_df$Genre)
spotify_artists_df$Genre <- ifelse(grepl("country", spotify_artists_df$Genre), "country", spotify_artists_df$Genre)
spotify_artists_df$Genre <- ifelse(grepl("hop", spotify_artists_df$Genre), "hip hop", spotify_artists_df$Genre)
spotify_artists_df$Genre <- ifelse(grepl("pop", spotify_artists_df$Genre), "pop", spotify_artists_df$Genre)
spotify_artists_df$Genre <- ifelse(grepl("rock", spotify_artists_df$Genre), "rock", spotify_artists_df$Genre)
spotify_artists_df$Genre <- ifelse(grepl("punk", spotify_artists_df$Genre), "rock", spotify_artists_df$Genre)
spotify_artists_df$Genre <- ifelse(grepl("motown", spotify_artists_df$Genre), "soul", spotify_artists_df$Genre)

spotify_artists_df

# Get spotify artist dataframe to csv

write.csv(spotify_artists_df, "spotify_artists_df.csv", row.names = FALSE)

# Webscrape the billboard website to get weeks on chart for billboard artist 100

get_rank <- function(artist) {
  rD <- rsDriver(browser = c("firefox"), verbose = FALSE, port = netstat::free_port(random = TRUE), chromever = NULL)
  driver <- rD$client

  # artist should be all in lowercase and if there are spaces replace with a dash
  
  artist <- tolower(artist)
  artist <- gsub(" ", "-", artist)
  url <- paste0("https://www.billboard.com/artist/", artist, "/chart-history/ats/")
  driver$navigate(url)
  
  # Locate the reject button:
  reject_button <- driver$findElement(using = "xpath", 
                                      value = '//*[@id="onetrust-reject-all-handler"]')
  # Click on the button:
  reject_button$clickElement()
  
  artist_rank <- driver$findElements(using = "css", value = ".artist-chart-row-week-on-chart")
  
  # if rank is not found, return NA
  
  if (length(artist_rank) == 0) {
    return(NA)
  }
  
  # if there are more than one number returned, return  NA
  
  if (length(artist_rank) > 1) {
    return(NA)
  }
  
  artist_rank <- sapply(artist_rank, function(x) {x$getElementText()})
  
  artist_rank <- as.numeric(artist_rank)
  
  return(artist_rank)
  
}
# Scrape in instalments of 10 to avoid being blocked
billboard_list_one <- lapply(spotify_artists_df$Artist[1:10], get_rank)
billboard_list_two <- lapply(spotify_artists_df$Artist[11:20], get_rank)
billboard_list_three <- lapply(spotify_artists_df$Artist[21:30], get_rank)
billboard_list_four <- lapply(spotify_artists_df$Artist[31:40], get_rank)
billboard_list_five <- lapply(spotify_artists_df$Artist[41:50], get_rank)
billboard_list_six <- lapply(spotify_artists_df$Artist[51:60], get_rank)
billboard_list_seven <- lapply(spotify_artists_df$Artist[61:70], get_rank)
billboard_list_eight <- lapply(spotify_artists_df$Artist[71:80], get_rank)
billboard_list_nine <- lapply(spotify_artists_df$Artist[81:90], get_rank)
billboard_list_ten <- lapply(spotify_artists_df$Artist[91:100], get_rank)
# Join all the lists together
billboard_list <- c(billboard_list_one, billboard_list_two, billboard_list_three, billboard_list_four, billboard_list_five, billboard_list_six, billboard_list_seven, billboard_list_eight, billboard_list_nine, billboard_list_ten)

# Convert to a dataframe
billboard_df <- data.frame(matrix(unlist(billboard_list), nrow=length(billboard_list), byrow=T))
billboard_df

# Create new table with billboard data
billboard_artists_df <- spotify_artists_df

# add list with weeks on chart to billboard_artists_df (and not billboard_df)
billboard_artists_df$weeks_on_chart <- billboard_list

# change weeks_on_chart to numeric
billboard_artists_df$weeks_on_chart <- as.numeric(billboard_artists_df$weeks_on_chart)

# where NA, replace with 0
billboard_artists_df$weeks_on_chart[is.na(billboard_artists_df$weeks_on_chart)] <- 0

# change so Smokey Robinson weeks_on_chart is 1, Simon & Garfunkel weeks_on_chart is 1, Tupac weeks_on_chart is 6 and AC/DC weeks_on_chart is 190 (as these artists name are different in the billboard charts)
billboard_artists_df$weeks_on_chart[billboard_artists_df$Artist_Name == "Smokey Robinson and the Miracles"] <- 1

billboard_artists_df$weeks_on_chart[billboard_artists_df$Artist_Name == "Simon and Garfunkel"] <- 1

billboard_artists_df$weeks_on_chart[billboard_artists_df$Artist_Name == "Tupac Shakur"] <- 6

billboard_artists_df$weeks_on_chart[billboard_artists_df$Artist_Name == "AC/DC"] <- 190

# change so weeks on chart is called Billboard Popularity 
colnames(billboard_artists_df)[colnames(billboard_artists_df) == "weeks_on_chart"] <- "Billboard_Popularity"

# change Billboard Popularity so it is current Billboard Popularity / 495 * 100
billboard_artists_df$Billboard_Popularity <- billboard_artists_df$Billboard_Popularity / 495 * 100

# change so Billboard Popularity has no decimal places
billboard_artists_df$Billboard_Popularity <- round(billboard_artists_df$Billboard_Popularity, 0)

billboard_artists_df

# Get billboard artist dataframe to csv

write.csv(billboard_artists_df, "billboard_artists_df.csv", row.names = FALSE)

# Create new table from spotify_artists_df 

spotify_track_df <- billboard_artists_df

# Get the artist top track for each artist in the spotify_artists_df data frame
get_artist_top <- function(artist_id){
  # Define the Spotify API endpoint for getting information about an artist
  top_url <- paste0('https://api.spotify.com/v1/artists/', artist_id , '/top-tracks?market=US')
  
  # Set up the request with the access token
  top_response <- GET(top_url, add_headers(Authorization = HeaderValue))
  
  # Extract the artist information from the response
  top_info <- content(top_response)
  
  track_id <- top_info$tracks[[1]]$id
  track_name <- top_info$tracks[[1]]$name
  track_release <- top_info$tracks[[1]]$album$release_date

  
  # Create a list with the extracted information
  artist_data <- list(track_id = track_id, track_name = track_name, track_release = track_release)
  
  return(artist_data)
}

# Apply the function to the entire "Artist_Name" column in the data frame
result <- lapply(spotify_track_df$Spotify_Artist_ID, get_artist_top)

# Extract individual elements
spotify_track_df$track_id <- sapply(result, function(x) x$track_id)
spotify_track_df$track_name <- sapply(result, function(x) x$track_name)
spotify_track_df$track_release <- sapply(result, function(x) x$track_release)


spotify_track_df

# Create new column called decade which derives only year from track_release

spotify_track_df$decade <- substr(spotify_track_df$track_release, 1, 4)

# change decade to numeric

spotify_track_df$decade <- as.numeric(spotify_track_df$decade)

# simplify decade (eg. if decade is between 1960 and 1969 then 1960)

spotify_track_df$decade[spotify_track_df$decade >= 1950 & spotify_track_df$decade <= 1959] <- 1950

spotify_track_df$decade[spotify_track_df$decade >= 1960 & spotify_track_df$decade <= 1969] <- 1960

spotify_track_df$decade[spotify_track_df$decade >= 1970 & spotify_track_df$decade <= 1979] <- 1970

spotify_track_df$decade[spotify_track_df$decade >= 1980 & spotify_track_df$decade <= 1989] <- 1980

spotify_track_df$decade[spotify_track_df$decade >= 1990 & spotify_track_df$decade <= 1999] <- 1990

spotify_track_df$decade[spotify_track_df$decade >= 2000 & spotify_track_df$decade <= 2009] <- 2000

spotify_track_df$decade[spotify_track_df$decade >= 2010 & spotify_track_df$decade <= 2019] <- 2010

spotify_track_df$decade[spotify_track_df$decade >= 2020 & spotify_track_df$decade <= 2029] <- 2020

spotify_track_df


# Get top track dataframe to csv
write.csv(spotify_track_df, "spotify_track_df.csv", row.names = FALSE)
# Create new table from spotify_track_df

spotify_audio_df <- spotify_track_df

# Get audio features for each track 

# Get the artist top track audio features 
get_audio <- function(track_id){
  # Define the Spotify API endpoint for getting information about an artist
  audio_url <- paste0('https://api.spotify.com/v1/audio-features/', track_id)
  
  # Set up the request with the access token
  audio_response <- GET(audio_url, add_headers(Authorization = HeaderValue))
  
  # Extract the artist information from the response
  audio_info <- content(audio_response)
  
  track_danceability <- audio_info$danceability
  track_energy <- audio_info$energy
  track_valence <- audio_info$valence
  
  # Create a list with the extracted information
  audio_data <- list(track_id = track_id,
                     track_danceability = track_danceability,
                     track_energy = track_energy, 
                     track_valence = track_valence)
  
  return(audio_data)
}

# Apply the function to the entire "Artist_Name" column in the data frame
result <- lapply(spotify_audio_df$track_id, get_audio)

# Extract individual elements
spotify_audio_df$track_danceability <- sapply(result, function(x) x$track_danceability)
spotify_audio_df$track_energy <- sapply(result, function(x) x$track_energy)
spotify_audio_df$track_valence <- sapply(result, function(x) x$track_valence)

# change name to Danceability, Energy, Valence
colnames(spotify_audio_df)[colnames(spotify_audio_df) == "track_danceability"] <- "Danceability"
colnames(spotify_audio_df)[colnames(spotify_audio_df) == "track_energy"] <- "Energy"
colnames(spotify_audio_df)[colnames(spotify_audio_df) == "track_valence"] <- "Valence"

spotify_audio_df

# Get spotify audio dataframe to csv
write.csv(spotify_audio_df, "spotify_audio_df.csv", row.names = FALSE)
# add endurance index to spotify_audio_df
spotify_audio_df$Endurance_Index <- (spotify_audio_df$Spotify_Popularity + spotify_audio_df$Billboard_Popularity) / 2

final_df <- spotify_audio_df
# Write  final dataframe to csv
write.csv(final_df, "final_df.csv", row.names = FALSE)
# Import final_df.csv

final_df <- read.csv("final_df.csv", header = TRUE)

# Create table from final_df which only contains ranking, artist name, spotify popularity, billboard popularity and endurance index

playlist_table <- final_df[,c(1,2,4,6,14)]

# write spotify theme for chart which can be found in reactable documentation

library(reactable)
library(htmltools)

spotify_theme <- function() {
  search_icon <- function(fill = "none") {
    # Icon from https://boxicons.com
    svg <- sprintf('<svg xmlns="http://www.w3.org/2000/svg" width="24" height="24"><path fill="%s" d="M10 18c1.85 0 3.54-.64 4.9-1.69l4.4 4.4 1.4-1.42-4.39-4.4A8 8 0 102 10a8 8 0 008 8.01zm0-14a6 6 0 11-.01 12.01A6 6 0 0110 4z"/></svg>', fill)
    sprintf("url('data:image/svg+xml;charset=utf-8,%s')", URLencode(svg))
  }

  text_color <- "hsl(0, 0%, 95%)"
  text_color_light <- "hsl(0, 0%, 70%)"
  text_color_lighter <- "hsl(0, 0%, 55%)"
  bg_color <- "hsl(0, 0%, 10%)"

  reactableTheme(
    color = text_color,
    backgroundColor = bg_color,
    borderColor = "hsl(0, 0%, 16%)",
    borderWidth = "1px",
    highlightColor = "rgba(255, 255, 255, 0.1)",
    cellPadding = "10px 8px",
    style = list(
      fontFamily = "Work Sans, Helvetica Neue, Helvetica, Arial, sans-serif",
      fontSize = "0.875rem",
      "a" = list(
        color = text_color,
        textDecoration = "none",
        "&:hover, &:focus" = list(
          textDecoration = "underline",
          textDecorationThickness = "1px"
        )
      ),
      ".number" = list(
        color = text_color_light,
        fontFamily = "Source Code Pro, Consolas, Monaco, monospace"
      ),
      ".tag" = list(
        padding = "0.125rem 0.25rem",
        color = "hsl(0, 0%, 40%)",
        fontSize = "0.75rem",
        border = "1px solid hsl(0, 0%, 24%)",
        borderRadius = "2px",
        textTransform = "uppercase"
      )
    ),
    headerStyle = list(
      color = text_color_light,
      fontWeight = 400,
      fontSize = "0.75rem",
      letterSpacing = "1px",
      textTransform = "uppercase",
      "&:hover, &:focus" = list(color = text_color)
    ),
    rowHighlightStyle = list(
      ".tag" = list(color = text_color, borderColor = text_color_lighter)
    ),
    # Full-width search bar with search icon
    searchInputStyle = list(
      paddingLeft = "1.9rem",
      paddingTop = "0.5rem",
      paddingBottom = "0.5rem",
      width = "100%",
      border = "none",
      backgroundColor = bg_color,
      backgroundImage = search_icon(text_color_light),
      backgroundSize = "1rem",
      backgroundPosition = "left 0.5rem center",
      backgroundRepeat = "no-repeat",
      "&:focus" = list(backgroundColor = "rgba(255, 255, 255, 0.1)", border = "none"),
      "&:hover, &:focus" = list(backgroundImage = search_icon(text_color)),
      "::placeholder" = list(color = text_color_lighter),
      "&:hover::placeholder, &:focus::placeholder" = list(color = text_color)
    ),
    paginationStyle = list(color = text_color_light),
    pageButtonHoverStyle = list(backgroundColor = "hsl(0, 0%, 20%)"),
    pageButtonActiveStyle = list(backgroundColor = "hsl(0, 0%, 24%)")
  )
}

# Function to create the chart 

rolling_stone_chart <- function(data) {
  reactable(
    data,
    theme = spotify_theme(),
    searchable = TRUE,
    defaultColDef = colDef(
      minWidth = 100,
      headerStyle = list(fontWeight = "bold")
    ),
    columns = list(
      Ranking = colDef(
        name = "#",
        align = "center",
        style = list(fontWeight = "bold")
      ),
      Artist_Name = colDef(
        name = "Artist",
        minWidth = 200,
        maxWidth = 400,
      ),
      Spotify_Popularity = colDef(
        name = "Spotify Popularity",
        align = "center",
        class = "number",
        style = list(fontWeight = "bold")
      ),
      Billboard_Popularity = colDef(
        name = "Billboard Popularity",
        align = "center",
        class = "number",
        style = list(fontWeight = "bold")
      ),
      Endurance_Index = colDef(
        name = "Endurance Index",
        align = "center",
        class = "number",
        style = list(fontWeight = "bold")
      )
    )
  )
}

div(
  rolling_stone_chart(playlist_table)
)
#set all figures as 9 by 6 inches
knitr::opts_chunk$set(fig.width=9, fig.height=6) 
library(plotly)

# Create figure one using plotly

# find mean current popularity
mean_current_popularity <- mean(final_df$Spotify_Popularity)


fig_one <- plot_ly(
  final_df, x = ~Spotify_Popularity, y = ~Ranking,
  color = ~Spotify_Popularity, size = ~Ranking, text = ~paste("Artist: ", Artist_Name, "<br>Ranking: ", Ranking, "<br>Current Popularity: ", Spotify_Popularity)
) %>%
  layout(
    title = '<b>Figure 1: Current Popularity vs Rolling Stone Ranking<b>',
    xaxis = list(title = "Current Popularity (Spotify)"),
    yaxis = list(title = "Rolling Stone Ranking"),
    shapes = list(
      list(
        type = "line",
        x0 = 0,
        y0 = 50,
        x1 = 100,
        y1 = 50,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = 50,
        y0 = 0,
        x1 = 50,
        y1 = 100,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = mean_current_popularity,
        y0 = 0,
        x1 = mean_current_popularity,
        y1 = 100,
        line = list(color = "#e7a4b6", width = 2)
      )
    )
  )

# add label at ranking = 51 and popularity = 1 as "Least Enduring", label at ranking = 1 and popularity = 100 as "Most Enduring", and label at ranking = 99 and popularity = 51 as "Highest Ranking" and label at ranking = 1 and popularity = 51 as "Lowest Ranking"

fig_one <- fig_one %>% layout(annotations = list(
  list(
    x = 92,
    y = 52,
    xref = "x",
    yref = "y",
    text = "Most Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 10,
    y = 52,
    xref = "x",
    yref = "y",
    text = "Least Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = -40,
    ay = 0
  ),
  list(
    x = 51,
    y = 102,
    xref = "x",
    yref = "y",
    text = "Highest Ranking",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 51,
    y = -2,
    xref = "x",
    yref = "y",
    text = "Lowest Ranking",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  ),
  list(
    x = mean_current_popularity,
    y = -2,
    xref = "x",
    yref = "y",
    text = "66",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  )
))


# add correlation coefficient to figure

correlation_coefficient_one <- cor(final_df$Ranking, final_df$Spotify_Popularity)


fig_one <- fig_one %>% layout(annotations = list(
  list(
    x = 90,
    y = 102,
    xref = "x",
    yref = "y",
    text = paste("Correlation Coefficient: ", round(correlation_coefficient_one, 2)),
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  )
))

fig_one

# find mean historical popularity
mean_historical_popularity <- mean(final_df$Billboard_Popularity)

fig_two <- plot_ly(
  final_df, x = ~Billboard_Popularity, y = ~Ranking,
  color = ~Billboard_Popularity, size = ~Ranking, text = ~paste("Artist: ", Artist_Name, "<br>Ranking: ", Ranking, "<br>Historical Popularity: ", Billboard_Popularity)
) %>%
  layout(
    title = '<b>Figure 2: Historical Popularity vs Rolling Stone Ranking<b>',
    xaxis = list(title = "Historical Popularity (Billboard)"),
    yaxis = list(title = "Rolling Stone Ranking"),
    shapes = list(
      list(
        type = "line",
        x0 = 0,
        y0 = 50,
        x1 = 100,
        y1 = 50,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = 50,
        y0 = 0,
        x1 = 50,
        y1 = 100,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = mean_historical_popularity,
        y0 = 0,
        x1 = mean_historical_popularity,
        y1 = 100,
        line = list(color = "#e7a4b6", width = 2)
      )
    )
  )

# add label at ranking = 51 and popularity = 1 as "Least Enduring", label at ranking = 1 and popularity = 100 as "Most Enduring", and label at ranking = 99 and popularity = 51 as "Highest Ranking" and label at ranking = 1 and popularity = 51 as "Lowest Ranking"

fig_two <- fig_two %>% layout(annotations = list(
  list(
    x = 92,
    y = 52,
    xref = "x",
    yref = "y",
    text = "Most Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 10,
    y = 52,
    xref = "x",
    yref = "y",
    text = "Least Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = -40,
    ay = 0
  ),
  list(
    x = 51,
    y = 102,
    xref = "x",
    yref = "y",
    text = "Highest Ranking",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 51,
    y = -2,
    xref = "x",
    yref = "y",
    text = "Lowest Ranking",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  ),
  list(
    x = mean_historical_popularity,
    y = -2,
    xref = "x",
    yref = "y",
    text = "8",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  )
))


# add correlation coefficient to figure

correlation_coefficient_two <- cor(final_df$Ranking, final_df$Billboard_Popularity)


fig_two <- fig_two %>% layout(annotations = list(
  list(
    x = 90,
    y = 102,
    xref = "x",
    yref = "y",
    text = paste("Correlation Coefficient: ", round(correlation_coefficient_two, 2)),
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  )
))


fig_two

# find mean endurance index
mean_endurance_index <- mean(final_df$Endurance_Index)

fig_three <- plot_ly(
  final_df, x = ~Endurance_Index, y = ~Ranking,
  color = ~Endurance_Index, size = ~Ranking, text = ~paste("Artist: ", Artist_Name, "<br>Ranking: ", Ranking, "<br>Endurance Index: ", Endurance_Index)
) %>%
  layout(
    title = '<b>Figure 3: Endurance Index vs Rolling Stone Ranking<b>',
    xaxis = list(title = "Endurance Index (Spotify & Billboard)"),
    yaxis = list(title = "Rolling Stone Ranking"),
    shapes = list(
      list(
        type = "line",
        x0 = 0,
        y0 = 50,
        x1 = 100,
        y1 = 50,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = 50,
        y0 = 0,
        x1 = 50,
        y1 = 100,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = mean_endurance_index,
        y0 = 0,
        x1 = mean_endurance_index,
        y1 = 100,
        line = list(color = "#e7a4b6", width = 2)
      )
    )
  )

# add label at ranking = 51 and popularity = 1 as "Least Enduring", label at ranking = 1 and popularity = 100 as "Most Enduring", and label at ranking = 99 and popularity = 51 as "Highest Ranking" and label at ranking = 1 and popularity = 51 as "Lowest Ranking"

fig_three <- fig_three %>% layout(annotations = list(
  list(
    x = 92,
    y = 52,
    xref = "x",
    yref = "y",
    text = "Most Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 10,
    y = 52,
    xref = "x",
    yref = "y",
    text = "Least Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = -40,
    ay = 0
  ),
  list(
    x = 51,
    y = 102,
    xref = "x",
    yref = "y",
    text = "Highest Ranking",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 51,
    y = -2,
    xref = "x",
    yref = "y",
    text = "Lowest Ranking",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  ),
    list(
    x = mean_endurance_index,
    y = -2,
    xref = "x",
    yref = "y",
    text = "37",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  )
))


# add correlation coefficient to figure

correlation_coefficient_three <- cor(final_df$Ranking, final_df$Endurance_Index)


fig_three <- fig_three %>% layout(annotations = list(
  list(
    x = 90,
    y = 102,
    xref = "x",
    yref = "y",
    text = paste("Correlation Coefficient: ", round(correlation_coefficient_three, 2)),
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  )
))


fig_three

# Create table with plotly which shows the relationship between genre and endurance index

# find mean endurance index for each genre
genre_endurance_index <- final_df %>%
  group_by(Genre) %>%
  summarise(Endurance_Index = mean(Endurance_Index))

# round endurance index to 0 decimal places
genre_endurance_index$Endurance_Index <- round(genre_endurance_index$Endurance_Index, 0)

fig_four <- plot_ly(
  genre_endurance_index, x = ~Genre, y = ~Endurance_Index,
  color = ~Genre, size = ~Endurance_Index, mode = "markers", type = "bar", text = ~paste("Genre: ", Genre, "<br>Endurance Index: ", Endurance_Index)
) %>%
  layout(
    title = '<b>Figure 4: Endurance Index vs Genre<b>',
    xaxis = list(title = "Genre"),
    yaxis = list(title = "Mean Endurance Index (Spotify & Billboard)")
  )

# add line at mean endurance index and there where y is approximately 37

fig_four <- fig_four %>% layout(shapes = list(
  list(
    type = "line",
    x0 = -0.5,
    y0 = 37,
    x1 = 7.5,
    y1 = 37,
    line = list(color = "#e7a4b6", width = 2)
  )
))

# add label saying 37 at y = 37

fig_four <- fig_four %>% layout(annotations = list(
  list(
    x = -0.7,
    y = 37,
    xref = "x",
    yref = "y",
    text = "37",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  )
))


fig_four
# Create table using plotly which shows the relationship between decade and endurance index

# find mean endurance index for each genre
decade_endurance_index <- final_df %>%
  group_by(decade) %>%
  summarise(Endurance_Index = mean(Endurance_Index))

# round endurance index to 0 decimal places
decade_endurance_index$Endurance_Index <- round(decade_endurance_index$Endurance_Index, 0)

fig_five <- plot_ly(
  decade_endurance_index, x = ~decade, y = ~Endurance_Index,
  color = ~decade, size = ~Endurance_Index, mode = "markers", type = "bar"
) %>%
  layout(
    title = '<b>Figure 5: Endurance Index vs Decade<b>',
    xaxis = list(title = "Decade"),
    yaxis = list(title = "Mean Endurance Index (Spotify & Billboard)")
  )

# add line at mean endurance index and there where y is approximately 37

fig_five <- fig_five %>% layout(shapes = list(
  list(
    type = "line",
    x0 = 1945,
    y0 = 37,
    x1 = 2025,
    y1 = 37,
    line = list(color = "#e7a4b6", width = 2)
  )
))

# add to graph correlation coefficient between decade and endurance index

correlation_coefficient_five <- cor(final_df$decade, final_df$Endurance_Index)

fig_five <- fig_five %>% layout(annotations = list(
  list(
    x = 2020,
    y = 46,
    xref = "x",
    yref = "y",
    text = paste("Correlation Coefficient: ", round(correlation_coefficient_five, 2)),
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  )
))

fig_five

# Create table using plotly which shows the relationship between danceability and endurance index

fig_six <- plot_ly(
  final_df, x = ~Endurance_Index, y = ~Danceability,
  color = ~Endurance_Index, size = ~Danceability, text = ~paste("Artist: ", Artist_Name, "Track: ", track_name, "<br>Ranking: ", Ranking, "<br>Endurance Index: ", Endurance_Index, "<br>Danceability: ", Danceability)
) %>%
  layout(
    title = '<b>Figure 6 : Endurance Index vs Danceability<b>',
    xaxis = list(title = "Endurance Index (Spotify & Billboard)"),
    yaxis = list(title = "Danceability (Spotify)"),
    shapes = list(
      list(
        type = "line",
        x0 = 0,
        y0 = 0.5,
        x1 = 100,
        y1 = 0.5,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = 50,
        y0 = 0,
        x1 = 50,
        y1 = 1,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = mean_endurance_index,
        y0 = 0,
        x1 = mean_endurance_index,
        y1 = 1,
        line = list(color = "#e7a4b6", width = 2)
      )
    )
  )

# add label at ranking = 51 and popularity = 1 as "Least Enduring", label at ranking = 1 and popularity = 100 as "Most Enduring", and label at ranking = 99 and popularity = 51 as "Most Danceable" and label at ranking = 1 and popularity = 51 as "Least Danceable"

fig_six <- fig_six %>% layout(annotations = list(
  list(
    x = 92,
    y = 0.52,
    xref = "x",
    yref = "y",
    text = "Most Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 10,
    y = 0.52,
    xref = "x",
    yref = "y",
    text = "Least Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = -40,
    ay = 0
  ),
  list(
    x = 51,
    y = 1.02,
    xref = "x",
    yref = "y",
    text = "Most Danceable",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 51,
    y = -0.02,
    xref = "x",
    yref = "y",
    text = "Least Danceable",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  ),
      list(
    x = mean_endurance_index,
    y = -0.02,
    xref = "x",
    yref = "y",
    text = "37",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  )
))


# add correlation coefficient to figure

correlation_coefficient_six <- cor(final_df$Danceability, final_df$Endurance_Index)


fig_six <- fig_six %>% layout(annotations = list(
  list(
    x = 90,
    y = 1.02,
    xref = "x",
    yref = "y",
    text = paste("Correlation Coefficient: ", round(correlation_coefficient_six, 2)),
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  )
))

# add line at mean endurance index


fig_six <- fig_six %>% layout(shapes = list(
  list(
    type = "line",
    x0 = mean_endurance_index,
    y0 = 0,
    x1 = mean_endurance_index,
    y1 = 1,
    line = list(color = "grey", width = 2)
  )
))

fig_six

# Create table using plotly which shows the relationship between energy and endurance index

fig_seven <- plot_ly(
  final_df, x = ~Endurance_Index, y = ~Energy,
  color = ~Endurance_Index, size = ~Energy, text = ~paste("Artist: ", Artist_Name, "Track: ", track_name, "<br>Ranking: ", Ranking, "<br>Endurance Index: ", Endurance_Index, "<br>Energy: ", Energy)
) %>%
  layout(
    title = '<b>Figure 7 : Endurance Index vs Energy<b>',
    xaxis = list(title = "Endurance Index (Spotify & Billboard)"),
    yaxis = list(title = "Energy (Spotify)"),
    shapes = list(
      list(
        type = "line",
        x0 = 0,
        y0 = 0.50,
        x1 = 100,
        y1 = 0.50,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = 50,
        y0 = 0,
        x1 = 50,
        y1 = 1,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = mean_endurance_index,
        y0 = 0,
        x1 = mean_endurance_index,
        y1 = 1,
        line = list(color = "#e7a4b6", width = 2)
      )
    )
  )

# add label at ranking = 51 and popularity = 1 as "Least Enduring", label at ranking = 1 and popularity = 100 as "Most Enduring", and label at ranking = 99 and popularity = 51 as "High Energy" and label at ranking = 1 and popularity = 51 as "Low Energy"

fig_seven <- fig_seven %>% layout(annotations = list(
  list(
    x = 92,
    y = 0.52,
    xref = "x",
    yref = "y",
    text = "Most Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 10,
    y = 0.52,
    xref = "x",
    yref = "y",
    text = "Least Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = -40,
    ay = 0
  ),
  list(
    x = 51,
    y = 1.02,
    xref = "x",
    yref = "y",
    text = "High Energy",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 51,
    y = -0.02,
    xref = "x",
    yref = "y",
    text = "Low Energy",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  ),
  list(
    x = mean_endurance_index,
    y = -0.02,
    xref = "x",
    yref = "y",
    text = "37",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = mean_endurance_index,
    y = -0.02,
    xref = "x",
    yref = "y",
    text = "37",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  )
))


# add correlation coefficient to figure

correlation_coefficient_seven <- cor(final_df$Energy, final_df$Endurance_Index)


fig_seven <- fig_seven %>% layout(annotations = list(
  list(
    x = 90,
    y = 1.02,
    xref = "x",
    yref = "y",
    text = paste("Correlation Coefficient: ", round(correlation_coefficient_seven, 2)),
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  )
))

# add line at mean endurance index

fig_seven <- fig_seven %>% layout(shapes = list(
  list(
    type = "line",
    x0 = mean_endurance_index,
    y0 = 0,
    x1 = mean_endurance_index,
    y1 = 1,
    line = list(color = "grey", width = 2)
  )
))

fig_seven

# Create table using plotly which shows the relationship between happiness (valence) and endurance index

fig_eight <- plot_ly(
  final_df, x = ~Endurance_Index, y = ~Valence,
  color = ~Endurance_Index, size = ~Valence, text = ~paste("Artist: ", Artist_Name, "Track: ", track_name, "<br>Ranking: ", Ranking, "<br>Endurance Index: ", Endurance_Index, "<br>Happiness: ", Valence)
) %>%
  layout(
    title = '<b>Figure 8 : Endurance Index vs Happiness<b>',
    xaxis = list(title = "Endurance Index (Spotify & Billboard)"),
    yaxis = list(title = "Happiness (Valence - Spotify)"),
    shapes = list(
      list(
        type = "line",
        x0 = 0,
        y0 = 0.50,
        x1 = 100,
        y1 = 0.50,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = 50,
        y0 = 0,
        x1 = 50,
        y1 = 1,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = mean_endurance_index,
        y0 = 0,
        x1 = mean_endurance_index,
        y1 = 1,
        line = list(color = "#e7a4b6", width = 2)
      )
    )
  )

# add label at ranking = 51 and popularity = 1 as "Least Enduring", label at ranking = 1 and popularity = 100 as "Most Enduring", and label at ranking = 99 and popularity = 51 as "Happier" and label at ranking = 1 and popularity = 51 as "Sadder"

fig_eight <- fig_eight %>% layout(annotations = list(
  list(
    x = 92,
    y = 0.52,
    xref = "x",
    yref = "y",
    text = "Most Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 10,
    y = 0.52,
    xref = "x",
    yref = "y",
    text = "Least Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = -40,
    ay = 0
  ),
  list(
    x = 51,
    y = 1.02,
    xref = "x",
    yref = "y",
    text = "Happier",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 51,
    y = -0.02,
    xref = "x",
    yref = "y",
    text = "Sadder",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  ),
  list(
    x = mean_endurance_index,
    y = -0.05,
    xref = "x",
    yref = "y",
    text = "37",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  )
))


# add correlation coefficient to figure

correlation_coefficient_eight <- cor(final_df$Valence, final_df$Endurance_Index)


fig_eight <- fig_eight %>% layout(annotations = list(
  list(
    x = 90,
    y = 1.02,
    xref = "x",
    yref = "y",
    text = paste("Correlation Coefficient: ", round(correlation_coefficient_eight, 2)),
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  )
))

# add line at mean endurance index

fig_eight <- fig_eight %>% layout(shapes = list(
  list(
    type = "line",
    x0 = mean_endurance_index,
    y0 = 0,
    x1 = mean_endurance_index,
    y1 = 1,
    line = list(color = "grey", width = 2)
  )
))

fig_eight
# create mood variable which is average of valence and energy and danceability

final_df <- final_df %>% mutate(Mood = (Valence + Energy + Danceability)/3)

# create table using plotly which shows the relationship between mood and endurance index

fig_nine <- plot_ly(
  final_df, x = ~Endurance_Index, y = ~Mood,
  color = ~Endurance_Index, size = ~Mood, text = ~paste("Artist: ", Artist_Name, "Track: ", track_name, "<br>Ranking: ", Ranking, "<br>Endurance Index: ", Endurance_Index, "<br>Mood: ", Mood)
) %>%
  layout(
    title = '<b>Figure 9 : Endurance Index vs Mood<b>',
    xaxis = list(title = "Endurance Index (Spotify & Billboard)"),
    yaxis = list(title = "Mood (Valence + Energy + Danceability - Spotify)"),
    shapes = list(
      list(
        type = "line",
        x0 = 0,
        y0 = 0.50,
        x1 = 100,
        y1 = 0.50,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = 50,
        y0 = 0,
        x1 = 50,
        y1 = 1,
        line = list(color = "grey", width = 2)
      ),
      list(
        type = "line",
        x0 = mean_endurance_index,
        y0 = 0,
        x1 = mean_endurance_index,
        y1 = 1,
        line = list(color = "#e7a4b6", width = 2)
      )
    )
  )

# add label at ranking = 51 and popularity = 1 as "Least Enduring", label at ranking = 1 and popularity = 100 as "Most Enduring", and label at ranking = 99 and popularity = 51 as "Musically Positive" and label at ranking = 1 and popularity = 51 as "Musically Negative"

fig_nine <- fig_nine %>% layout(annotations = list(
  list(
    x = 92,
    y = 0.52,
    xref = "x",
    yref = "y",
    text = "Most Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 10,
    y = 0.52,
    xref = "x",
    yref = "y",
    text = "Least Enduring",
    showarrow = FALSE,
    arrowhead = 0,
    ax = -40,
    ay = 0
  ),
  list(
    x = 51,
    y = 1.02,
    xref = "x",
    yref = "y",
    text = "Musically Positive (Danceable, Energetic & Happy)",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  ),
  list(
    x = 51,
    y = -0.02,
    xref = "x",
    yref = "y",
    text = "Musically Negative (Slow, Calm & Sad)",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  ),
  list(
    x = mean_endurance_index,
    y = -0.05,
    xref = "x",
    yref = "y",
    text = "37",
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = -40
  )
))


# add correlation coefficient to figure

correlation_coefficient_nine <- cor(final_df$Mood, final_df$Endurance_Index)


fig_nine <- fig_nine %>% layout(annotations = list(
  list(
    x = 90,
    y = 0.98,
    xref = "x",
    yref = "y",
    text = paste("Correlation Coefficient: ", round(correlation_coefficient_nine, 2)),
    showarrow = FALSE,
    arrowhead = 0,
    ax = 0,
    ay = 40
  )
))

# add line at mean endurance index

fig_nine <- fig_nine %>% layout(shapes = list(
  list(
    type = "line",
    x0 = mean_endurance_index,
    y0 = 0,
    x1 = mean_endurance_index,
    y1 = 1,
    line = list(color = "grey", width = 2)
  )
))

fig_nine